aclnnSmoothL1LossBackward

📄 View Source Code

Applicable Products

Function

• Description: Computes the backward propagation of the aclnnSmoothL1Loss API.

• Formula:

  The backward propagation of SmoothL1Loss can be computed by derivation. For the first case of SmoothL1Loss, i.e., |x-y|<1, its derivative is:

  $$\frac{\partial SmoothL1Loss(x,y)}{\partial x} = x - y$$

  For the second case of SmoothL1Loss, where |x-y|≥1, the derivative is:

  $$\frac{\partial SmoothL1Loss(x,y)}{\partial x} = sign(x - y)$$

  where sign(x) denotes the sign function of x, that is:

  $$sign(x) =\begin{cases} 1,& if　x>0 \\ 0,& if　x=0 \\ -1,& if　x<0 \end{cases}$$

Function Prototype

Each operator is divided into a two-phase API. First call the "aclnnSmoothL1LossBackwardGetWorkspaceSize" API to obtain the workspace size required for computation and the executor that contains the operator computation process, and then call the "aclnnSmoothL1LossBackward" API to perform the computation.

aclnnStatus aclnnSmoothL1LossBackwardGetWorkspaceSize(
    const aclTensor* gradOutput,
    const aclTensor* self,
    const aclTensor* target,
    int64_t          reduction,
    float            beta,
    aclTensor*       gradInput,
    uint64_t*        workspaceSize,
    aclOpExecutor**  executor)

aclnnStatus aclnnSmoothL1LossBackward(
    void            *workspace,
    uint64_t         workspaceSize,
    aclOpExecutor   *executor,
    aclrtStream      stream)

aclnnSmoothL1LossBackwardGetWorkspaceSize

• Parameters:

  Parameter	Input/Output	Description	Instruction	Data Type	Data Format	Dimension (Shape)	Non-Contiguous Tensor
  gradOutput (aclTensor*)	Input	Gradient backward input, which is SmoothL1Loss in the formula.	The shape must satisfy the broadcast relationship with self and target. The data type must satisfy the data type derivation rule with the data types of self and target.	FLOAT, FLOAT16, BFLOAT16	ND	1-8	√
  self (aclTensor*)	Input	Input tensor, representing the input x in the formula.	The shape must satisfy the broadcast relationship with gradOutput and target. The data type must comply with the data type derivation rule with gradOutput and target.	FLOAT, FLOAT16, BFLOAT16	ND	1-8	√
  target (aclTensor*)	Input	The ground truth labels, i.e., input y in the formula.	The shape must satisfy the broadcast relationship with gradOutput and self. The data types of gradOutput and target must satisfy the data type derivation rule.	FLOAT, FLOAT16, BFLOAT16	ND	1-8	√
  reduction（int64_t）	Input	Specifies the reduction to apply to the output, i.e., the parameter reduction in the formula.	Supports 0 (none) | 1 (mean) | 2 (sum). 'none' indicates no reduction is applied. 'mean' indicates the sum of the output is divided by the number of elements in the output. 'sum' indicates the output is summed.	INT64	-	-	√
  beta (float)	Input	Computation attribute that specifies the threshold at which the loss function switches between L1 and L2 loss.	This value must be non-negative.	FLOAT	-	-	-
  gradInput (aclTensor*)	Output	Computed output.	Shape is the broadcast result of gradOutput, self, and target.	FLOAT, FLOAT16, BFLOAT16	ND	1-8	√
  workspaceSize (uint64_t*)	Output	Returns the workspace size to be requested on the Device side.	-	-	-	-	-
  executor (aclOpExecutor**)	Output	Returns the op executor, which contains the operator computation flow.	-	-	-	-	-

• Return value

  aclnnStatus: return code. For details, see aclnn Return Code.

  The first-phase API performs input parameter validation and returns an error in the following scenarios:

  Return Value	Return Code	Description
  ACLNN_ERR_PARAM_NULLPTR	161001	The passed self, target, gradOutput, or gradInput is a null pointer.
  ACLNN_ERR_PARAM_INVALID	161002	The data type of self, target, gradOutput, or gradInput is not within the supported range.
  		The shape of self, target, gradOutput, or gradInput does not meet the constraints.
  		The reduction does not meet the constraints.
  		The shapes of self, target, and gradOutput do not meet the requirements in Parameters.

aclnnSmoothL1LossBackward

• Parameters:

  Parameter	Input/Output	Description
  workspace	Input	Address of the workspace memory applied for on the Device side.
  workspaceSize	Input	Size of the workspace applied for on the Device side, obtained from the first-phase API aclnnSmoothL1LossBackwardGetWorkspaceSize.
  executor	Input	Operator executor, which contains the operator computation flow.
  stream	Input	Specifies the stream for task execution.

• Return value

  aclnnStatus: return code. For details, see aclnn Return Code.

Constraints

• Deterministic computation

  • aclnnSmoothL1LossBackward defaults to a deterministic implementation.

Example

The following provides example code for reference only. For details about compilation and running, see Compile and Run Samples.

#include <iostream>
#include <vector>
#include "acl/acl.h"
#include "aclnnop/aclnn_smooth_l1_loss_backward.h"

#define CHECK_RET(cond, return_expr) \
  do {                               \
    if (!(cond)) {                   \
      return_expr;                   \
    }                                \
  } while (0)

#define LOG_PRINT(message, ...)     \
  do {                              \
    printf(message, ##__VA_ARGS__); \
  } while (0)

int64_t GetShapeSize(const std::vector<int64_t>& shape) {
  int64_t shapeSize = 1;
  for (auto i : shape) {
    shapeSize *= i;
  }
  return shapeSize;
}

int Init(int32_t deviceId, aclrtStream* stream) {
  // Boilerplate: resource initialization.
  auto ret = aclInit(nullptr);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclInit failed. ERROR: %d\n", ret); return ret);
  ret = aclrtSetDevice(deviceId);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtSetDevice failed. ERROR: %d\n", ret); return ret);
  ret = aclrtCreateStream(stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtCreateStream failed. ERROR: %d\n", ret); return ret);
  return 0;
}

template <typename T>
int CreateAclTensor(const std::vector<T>& hostData, const std::vector<int64_t>& shape, void** deviceAddr,
                    aclDataType dataType, aclTensor** tensor) {
  auto size = GetShapeSize(shape) * sizeof(T);
  // Call aclrtMalloc to apply for device-side memory.
  auto ret = aclrtMalloc(deviceAddr, size, ACL_MEM_MALLOC_HUGE_FIRST);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtMalloc failed. ERROR: %d\n", ret); return ret);
  // Copy data from the host to the device memory by calling aclrtMemcpy.
  ret = aclrtMemcpy(*deviceAddr, size, hostData.data(), size, ACL_MEMCPY_HOST_TO_DEVICE);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtMemcpy failed. ERROR: %d\n", ret); return ret);

  // Calculate the strides of a contiguous tensor.
  std::vector<int64_t> strides(shape.size(), 1);
  for (int64_t i = shape.size() - 2; i >= 0; i--) {
    strides[i] = shape[i + 1] * strides[i + 1];
  }

  // Call the aclCreateTensor API to create an aclTensor.
  *tensor = aclCreateTensor(shape.data(), shape.size(), dataType, strides.data(), 0, aclFormat::ACL_FORMAT_ND,
                            shape.data(), shape.size(), *deviceAddr);
  return 0;
}

int main() {
  // 1. (Boilerplate) Initialize the device/stream. See the ACL API manual.
  // Set deviceId based on your actual device.
  int32_t deviceId = 0;
  aclrtStream stream;
  auto ret = Init(deviceId, &stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("Init acl failed. ERROR: %d\n", ret); return ret);

  // 2. Construct the input and output. Customize the construction based on the API.
  std::vector<int64_t> gradOutputShape = {4, 2};
  std::vector<int64_t> selfShape = {4, 2};
  std::vector<int64_t> targetShape = {4, 2};
  std::vector<int64_t> gradInputShape = {4, 2};
  int64_t reduction = 0;
  float beta = 1.0;
  void* gradOutputDeviceAddr = nullptr;
  void* selfDeviceAddr = nullptr;
  void* targetDeviceAddr = nullptr;
  void* gradInputDeviceAddr = nullptr;
  aclTensor* gradOutput = nullptr;
  aclTensor* self = nullptr;
  aclTensor* target = nullptr;
  aclTensor* gradInput = nullptr;
  std::vector<float> gradOutputHostData = {1, 1, 1, 1, 1, 1, 1, 1};
  std::vector<float> selfHostData = {0, 1, 2, 3, 4, 5, 6, 7};
  std::vector<float> targetHostData = {1, 1, 1, 1, 1, 1, 1, 1};
  std::vector<float> gradInputHostData(8, 0);
  // Create the gradOutput aclTensor.
  ret = CreateAclTensor(gradOutputHostData, gradOutputShape, &gradOutputDeviceAddr, aclDataType::ACL_FLOAT, &gradOutput);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // Create the self aclTensor.
  ret = CreateAclTensor(selfHostData, selfShape, &selfDeviceAddr, aclDataType::ACL_FLOAT, &self);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // Create the target aclTensor.
  ret = CreateAclTensor(targetHostData, targetShape, &targetDeviceAddr, aclDataType::ACL_FLOAT, &target);
  CHECK_RET(ret == ACL_SUCCESS, return ret);
  // Create the gradInput aclTensor.
  ret = CreateAclTensor(gradInputHostData, gradInputShape, &gradInputDeviceAddr, aclDataType::ACL_FLOAT, &gradInput);
  CHECK_RET(ret == ACL_SUCCESS, return ret);

  // 3. Call the CANN operator library API. Replace with the actual API name.
  uint64_t workspaceSize = 0;
  aclOpExecutor* executor;
  // Call the first-phase API of aclnnSmoothL1LossBackward.
  ret = aclnnSmoothL1LossBackwardGetWorkspaceSize(gradOutput, self, target, reduction, beta, gradInput, &workspaceSize, &executor);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnSmoothL1LossBackwardGetWorkspaceSize failed. ERROR: %d\n", ret); return ret);
  // Apply for device memory based on the workspaceSize computed by the first-phase API.
  void* workspaceAddr = nullptr;
  if (workspaceSize > 0) {
    ret = aclrtMalloc(&workspaceAddr, workspaceSize, ACL_MEM_MALLOC_HUGE_FIRST);
    CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("allocate workspace failed. ERROR: %d\n", ret); return ret);
  }
  // Call the second-phase API of aclnnSmoothL1LossBackward.
  ret = aclnnSmoothL1LossBackward(workspaceAddr, workspaceSize, executor, stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclnnSmoothL1LossBackward failed. ERROR: %d\n", ret); return ret);

  // 4. (Boilerplate) Wait synchronously for task execution to complete.
  ret = aclrtSynchronizeStream(stream);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("aclrtSynchronizeStream failed. ERROR: %d\n", ret); return ret);

  // 5. Obtain the output value and copy the result from the device-side memory to the host side. Modify this based on the specific API definition.
  auto size = GetShapeSize(gradInputShape);
  std::vector<float> resultData(size, 0);
  ret = aclrtMemcpy(resultData.data(), resultData.size() * sizeof(resultData[0]), gradInputDeviceAddr,
                    size * sizeof(resultData[0]), ACL_MEMCPY_DEVICE_TO_HOST);
  CHECK_RET(ret == ACL_SUCCESS, LOG_PRINT("copy result from device to host failed. ERROR: %d\n", ret); return ret);
  for (int64_t i = 0; i < size; i++) {
    LOG_PRINT("result[%ld] is: %f\n", i, resultData[i]);
  }

  // 6. Release aclTensor and aclScalar. Modify this based on the specific API definition.
  aclDestroyTensor(gradOutput);
  aclDestroyTensor(self);
  aclDestroyTensor(target);
  aclDestroyTensor(gradInput);
  // 7. Release device resources. Set the parameters based on the specific API definition.
  aclrtFree(gradOutputDeviceAddr);
  aclrtFree(selfDeviceAddr);
  aclrtFree(targetDeviceAddr);
  aclrtFree(gradInputDeviceAddr);
  if (workspaceSize > 0) {
    aclrtFree(workspaceAddr);
  }
  aclrtDestroyStream(stream);
  aclrtResetDevice(deviceId);
  aclFinalize();
  return 0;
}